Abstract
Background Despite its many known shortcomings, body mass index (BMI) is the most widely used measure of obesity, in part because of its practicality. Other more physiologic measurements of obesity have been proposed, including percent body fat (BF). Few studies have compared BMI and BF, especially in patients with coronary heart disease (CHD).
Methods We studied 581 patients with CHD following major CHD events. We divided patients into low (≤ 25% in men and ≤ 35% in women) and high BF (> 25% in men and > 35% in women) as determined by the sum of the skin-fold method and compared these findings with standard BMI determinations.
Results BMI and BF were highly correlated (r = 0.60; P < .001) and classified patients in the same category (obese vs nonobese) in 68% of cases. The agreement was optimal in the underweight (BMI < 18.5 kg/m2) and obese class II category (BMI 35-39.9 kg/m2), in which 100% of patients were classified as nonobese and obese, respectively, by both BMI and BF categories. The performance was worst in patients with BMIs in the overweight or preobese range (25-29.9 kg/m2), in which 58% of patients would be classified as obese according to BF criteria.
Conclusions Although some CHD patients are classified differently by BMI and BF, especially within the overweight BMI group, in general BMI and BF are highly correlated, especially in the underweight and obese BMI groups. Prospective studies are needed to determine which index of obesity best predicts risk in primary and secondary prevention.
INTRODUCTION
According to the World Health Organization (WHO),1 in 2005 approximately 1.6 billion adults were overweight and about 400 million were obese; the WHO projects that by 2015 2.3 billion adults will be overweight and more than 700 million will be obese. In the United States alone, it is estimated that nearly 68% of adults are overweight and 35% are obese. Overweight and obesity accounted for 9.1% of total US medical expenditures in 1998 and may have reached as high as $78.5 billion ($492.6 billion in 2002 dollars) with approximately half of this amount paid by Medicare.2,3
Obesity has been defined in terms of body mass index (BMI) when associated with various comorbidities: hypertension,4-⇓6 type 2 diabetes,7,8 coronary heart disease (CHD),9,10 stroke, gallbladder disease,11 and sleep apnea. On the other hand, WHO defines overweight and obesity in terms of fat: “abnormal or excessive fat accumulation that may impair health.” Although BMI is the most frequently used method to assess body composition, it does not always reflect body fatness, especially in certain populations.12,13 In fact, some studies14 suggest that BMI underestimates the prevalence of obesity. Because of differences in body composition, some experts have considered the percent body fat (BF) more appropriate in defining obesity,15 and BF > 25% in men or > 35% in women has been recommended for defining obesity in a major endocrinology society's guideline statement.16
Several studies have examined the relationship between BMI and body composition; Wellens et al17 found that BMI correctly identified about 44% of obese men and 52% of obese women when obesity was determined from BF (using densitometry) in a population of 504 men and 511 women ages 20-45 years. Frankenfield et al18 used bioelectrical impedance and found that all individuals determined to be obese by BMI were, in fact, obese by BF; however, 36% of men and 46% of women with BMI below 30 kg/m2 had obesity levels using BF, suggesting that BMI may not be an adequate assessment of body fatness.19-⇓⇓⇓⇓⇓⇓26
In the present study, we examine the concordance of BMI and BF in classifying overweightness/obesity in a population with stable CHD. In addition, we examine the correlation between BMI and BF, as well as the factors that influence this correlation.
METHODS
Patients
We retrospectively reviewed records of 581 patients who attended our cardiac rehabilitation and exercise training program following major CHD events. We gathered demographic, lipid, and clinical data as previously described.27-⇓⇓⇓⇓⇓⇓⇓⇓36 We divided patients into low BF or nonobese (≤ 25% men and ≤ 35% women) and high BF or obese (> 25% in men and > 35% in women) as described above. At the same time, we divided the population into 6 groups based on the WHO BMI thresholds. The goal was to examine the BF-BMI correlation and compare how the patients would be classified differently according to BMI and BF. The study was approved by the Institutional Review Board at Ochsner Clinic Foundation, New Orleans, LA.
Protocol
At baseline, fasting plasma lipids, glucose, high-sensitivity C-reactive protein, and percent BF were measured. All measurements were made in the early morning before exercise; BF was assessed by the sum of the skin-fold method using an average of 3 skin folds—thigh, chest, and abdomen in men; thigh, triceps, and supra-iliac in women.37 Cardiopulmonary exercise testing was performed as previously described.38,39 Subjects were divided into obese and nonobese according to BMI and according to BF.
Statistical Analysis
SAS 9.0 (SAS Institute, Cary, NC) was used for the statistical analysis. Results are mean ± standard deviation or frequencies expressed as percentages. Pearson's correlation was used to determine correlation, and a scatterplot was generated with 95% confidence ellipse.
RESULTS
Baseline Characteristics
We found 581 patients with both BMI and percent BF measured. Table 1 summarizes baseline characteristics of this study population. The group was 27% females and had a mean age of 64 ± 11 years. A history of hypertension was present in 34%, 24% had diabetes, and only 1.3% admitted to being a current smoker.
Baseline Characteristics of the Population
Correlations
Overall, we found a strong statistically significant correlation (r = 0.60; P < .001) between BMI and percent BF (Figure). The correlations in the subgroups were statistically significant (P < .0001) in older and younger patients, in both the obese and nonobese, and across both sexes (Table 2).
Scatter plot showing the correlation between body fat and body mass index.
Pearson Correlations for Population by Age (cutoff of 65) and Sex
Classifications
Overall, both classifications labeled subjects similarly in 68% of cases, 223 patients in the low BF group and 174 patients in the high BF group (Table 3). Using percent BF, 244 (42%) subjects were labeled as nonobese, 21 (9%) of whom were labeled as obese using BMI criteria. The other 337 (58%) subjects were labeled as obese according to BF, but 163 (48%) of these did not meet criteria for obesity according to BMI. The vast majority of the differently classified subjects—the 137 overweight patients (74%)—had a BMI of 25-30 kg/m2. In both the low BF and the high BF groups, 182 (99%) of misclassified subjects had BMI of 18.5-35 kg/m2. Two subjects had BMI ≥ 40 kg/m2 but were classified as nonobese by BF. BMI classified 195 (34%) subjects as obese, while BF classified 337 (58%) subjects as obese.
Agreement Between Established Categories of BMI and Percent BF
DISCUSSION
Correlation of BF and BMI
The correlation of BF and BMI seems to be dependent on multiple factors, including race, sex, age, and BMI itself.40-⇓42 In the Heritage Family Study, Jackson et al43 found that for similar BMIs, percent BF could vary as much as 10% between the sexes. Gallagher et al44 observed statistically significant age differences in all the studied groups. Movsesyan et al45 also saw this result in a study of Danish women.
The variations in the relationship between BMI and percent BF in different ethnic populations46 has led researchers to examine many different ethnic subgroups.47-⇓⇓⇓⇓⇓53 Some propose lower BMI cutoffs for Asian subgroups.14,54 The Heritage Family Study showed that the variation between races is larger for women than for men.43 BMI itself can also influence this relationship: In the severely obese, the gap between BMI and BF increases.22,23 These findings were also apparent in our study of patients with stable CHD (Table 3), although we did not have information regarding race.
BMI or BF in Defining Obesity
The BMI-based definition of obesity classified fewer people as obese than did the percent BF method, raising the question whether BMI or BF should be the gold standard for assessing body composition. Some ethnic populations have low BMIs and high BF and have a higher CHD prevalence than their low BF counterparts.55 Stroke has been associated with percent BF rather than BMI.56 Similarly, adiposity is associated with gallstone formation.57
Although we used a relatively simple and inexpensive assessment of BF (the sum of the skin-fold method), this method may not be as accurate as more sophisticated options. There is not a single, universally accepted method for assessing body composition.58 However, dual-energy x-ray absorptiometry (DEXA) has become one of the most widely used methods for estimating BF percentage because of its rapid, relatively easy operation and accuracy.59 Air displacement plethysmography,60-⇓62 bioelectrical impedance,63-⇓65 and hydrostatic weighing66,67 are other commonly used techniques. All of these methods, however, require significant financial and time investments, and none can easily be performed in the typical outpatient setting. These drawbacks explain why large epidemiologic studies still use BMI.
Different equations17,21,43,44,68,69 have been proposed to predict BF based on BMI in different populations—trying to account for age, sex, and ethnicity—with varying success.43,70 Such equations demonstrated the above-mentioned dependence of BMI and BF on multiple factors, especially age and sex, and they represent a relatively useful tool in classifying individuals as overweight and obese. Our study in CHD patients demonstrates that BMI and BF classifications are generally well correlated, especially with the extremes of BMI, although significant variability exists, especially in the middle BMI ranges.
Limitations
This study had multiple limitations. First, the study group was based on eligibility for cardiac rehabilitation and thus may not be representative. Second, our population had relatively high BMI and BF. Third, we measured BF using the skin-fold method, which is not considered to be the most accurate technique to determine BF.37 Last, our population mainly included older patients (mean age 64 ± 11 years), a group for which BMI and BF overlap to a lesser extent.20
CONCLUSIONS
Although BMI and BF correlate closely, their relationship depends on age, sex, ethnicity, and the exact level of BMI. BMI and BF often differ in classifying subjects as obese, with BMI tending to less often identify people as obese than the WHO percent BF classification. Further studies are needed to determine what body composition measurement best predicts risk in primary and secondary prevention of CHD.
This article meets the Accreditation Council for Graduate Medical Education competencies for Patient Care and Medical Knowledge.
- Academic Division of Ochsner Clinic Foundation
REFERENCES
- 1.↵FlegalKMCarrollMDOgdenCLCurtinLR (2010) Prevalence and trends in obesity among US adults, 1999-2008. JAMA 303(3):235–241, 20071471.
- 2.↵FinkelsteinEAFiebelkornICWangG (2003) National medical spending attributable to overweight and obesity: how much, and who's paying. Health Aff (Millwood) Suppl Web Exclusives:W3-219–W3-226, 14527256.
- 3.↵FinkelsteinEAFiebelkornICWangG (2004) State-level estimates of annual medical expenditures attributable to obesity. Obes Res 12(1):18–24, 14742838.
- 4.↵BrownCDHigginsMDonatoKA (2000) Body mass index and the prevalence of hypertension and dyslipidemia. Obes Res 8(9):605–619, 11225709.
- 5.↵StamlerRStamlerJRiedlingerWFAlgeraGRobertsRH (1978) Weight and blood pressure. Findings in hypertension screening of 1 million Americans. JAMA 240(15):1607–1610, 691146.
- 6.↵DyerARElliottP (1989) The INTERSALT study: relations of body mass index to blood pressure. INTERSALT Co-operative Research Group. J Hum Hypertens 3(5):299–308, 2810326.
- 7.↵ChanJMRimmEBColditzGAStampferMJWillettWC (1994) Obesity, fat distribution, and weight gain as risk factors for clinical diabetes in men. Diabetes Care 17(9):961–969, 7988316.
- 8.↵OhlsonLOLarssonBSvärdsuddK (1985) The influence of body fat distribution on the incidence of diabetes mellitus. 13.5 years of follow-up of the participants in the study of men born in 1913. Diabetes 34(10):1055–1058, 4043554.
- 9.↵Health implications of obesity (1985) National Institutes of Health Consensus Development Conference Statement. Ann Intern Med 103(6 Pt 2):1073–1077, 4062128.
- 10.↵HubertHBFeinleibMMcNamaraPMCastelliWP (1983) Obesity as an independent risk factor for cardiovascular disease: a 26-year follow-up of participants in the Framingham Heart Study. Circulation 67(5):968–977, 6219830.
- 11.↵StampferMJMaclureKMColditzGAMansonJEWillettWC (1992) Risk of symptomatic gallstones in women with severe obesity. Am J Clin Nutr 55(3):652–658, 1550039.
- 12.↵De LorenzoADeurenbergPPietrantuonoMDi DanieleNCervelliVAndreoliA (2003) How fat is obese. Acta Diabetol 40(Suppl 1):S254–S257, 14618486.
- 13.↵Deurenberg-YapMSchmidtGvan StaverenWADeurenbergP (2000) The paradox of low body mass index and high body fat percentage among Chinese, Malays and Indians in Singapore. Int J Obes Relat Metab Disord 24(8):1011–1017, 10951540.
- 14.↵LanhamDASteadMATsangKDaviesPS (2001) The prediction of body composition in Chinese Australian females. Int J Obes Relat Metab Disord 25(2):286–291, 11410833.
- 15.↵PoirierP (2007) Adiposity and cardiovascular disease: are we using the right definition of obesity. Eur Heart J 28(17):2047–2048, 17673449.
- 16.↵DickeyRABartuskaCDBrayGW (1998) AACE/ACE position statement on the prevention, diagnosis, and treatment of obesity (1998 revision). Endocr Pract 4(5):297–350.
- 17.↵WellensRIRocheAFKhamisHJJacksonASPollockMLSiervogelRM (1996) Relationships between the Body Mass Index and body composition. Obes Res 4(1):35–44, 8787936.
- 18.↵FrankenfieldDCRoweWACooneyRNSmithJSBeckerD (2001) Limits of body mass index to detect obesity and predict body composition. Nutrition 17(1):26–30, 11165884.
- 19.↵GallagherDHeymsfieldSBHeoMJebbSAMurgatroydPRSakamotoY (2000) Healthy percentage body fat ranges: an approach for developing guidelines based on body mass index. Am J Clin Nutr 72(3):694–701, 10966886.
- 20.↵DeurenbergPvan der KooyKHulshofTEversP (1989) Body mass index as a measure of body fatness in the elderly. Eur J Clin Nutr 43(4):231–236, 2661215.
- 21.↵DeurenbergPWeststrateJASeidellJC (1991) Body mass index as a measure of body fatness: age- and sex-specific prediction formulas. Br J Nutr 65(2):105–114, 2043597.
- 22.↵BaumgartnerRNHeymsfieldSBRocheAF (1995) Human body composition and the epidemiology of chronic disease. Obes Res 3(1):73–95, 7712363.
- 23.↵AdamsTDHeathEMLaMonteMJ (2007) The relationship between body mass index and per cent body fat in the severely obese. Diabetes Obes Metab 9(4):498–505, 17587392.
- 24.↵WangJGallagherDThorntonJC (2007) Regional body volumes, BMI, waist circumference, and percentage fat in severely obese adults. Obesity (Silver Spring) 15(11):2688–2698, 18070760.
- 25.↵Romero-CorralASomersVKSierra-JohnsonJ (2007) Diagnostic performance of body mass index to detect obesity in patients with coronary artery disease. Eur Heart J 28(17):2087–2093, 17626030.
- 26.↵Romero-CorralASomersVKSierra-JohnsonJ (2008) Accuracy of body mass index in diagnosing obesity in the adult general population. Int J Obes (Lond) 32(6):959–966, 18283284.
- 27.↵LavieCJMilaniRV (1994) Patients with high baseline exercise capacity benefit from cardiac rehabilitation and exercise training programs. Am Heart J 128(6 Pt 1):1105–1109, 7985590.
- 28.↵LuciniDMilaniRVCostantinoGLavieCJPortaAPaganiM (2002) Effects of cardiac rehabilitation and exercise training on autonomic regulation in patients with coronary artery disease. Am Heart J 143(6):977–983, 12075252.
- 29.↵LavieCJMilaniRV (1994) Effects of cardiac rehabilitation and exercise training on low-density lipoprotein cholesterol in patients with hypertriglyceridemia and coronary artery disease. Am J Cardiol 74(12):1192–1195, 7977088.
- 30.↵LavieCJMilaniRV (1997) Effects of cardiac rehabilitation, exercise training, and weight reduction on exercise capacity, coronary risk factors, behavioral characteristics, and quality of life in obese coronary patients. Am J Cardiol 79(4):397–401, 9052338.
- 31.↵LavieCJMilaniRV (2004) Prevalence of anxiety in coronary patients with improvement following cardiac rehabilitation and exercise training. Am J Cardiol 93(3):336–339, 14759385.
- 32.↵LavieCJMilaniRV (2005) Prevalence of hostility in young coronary artery disease patients and effects of cardiac rehabilitation and exercise training. Mayo Clin Proc 80(3):335–342, 15757014.
- 33.↵MilaniRVLavieCJMehraM. R (2004) Reduction in C-reactive protein through cardiac rehabilitation and exercise training. J Am Coll Cardiol 43(6):1056–1061, 15028366.
- 34.↵MilaniRVLavieCJ (2007) Impact of cardiac rehabilitation on depression and its associated mortality. Am J Med 120(9):799–806, 17765050.
- 35.↵LavieCJMilaniRV (2006) Adverse psychological and coronary risk profiles in young patients with coronary artery disease and benefits of formal cardiac rehabilitation. Arch Intern Med 166(17):1878–1883, 17000945.
- 36.↵ChurchTSLavieCJMilaniRVKirbyGS (2002) Improvements in blood rheology after cardiac rehabilitation and exercise training in patients with coronary heart disease. Am Heart J 143(2):349–355, 11835042.
- 37.↵HeywardVH (1998) Practical body composition assessment for children, adults, and older adults. Int J Sport Nutr 8(3):285–307, 9738136.
- 38.↵MilaniRVLavieCJMehraMR (2004) Cardiopulmonary exercise testing: how do we differentiate the cause of dyspnea. Circulation 110(4):e27–e31, 15277333.
- 39.↵MilaniRVLavieCJMehraMRVenturaHO (2006) Understanding the basics of cardiopulmonary exercise testing. Mayo Clin Proc 81(12):1603–1611, 17165639.
- 40.↵KyleUGSchutzYDupertuisYMPichardC (2003) Body composition interpretation. Contributions of the fat-free mass index and the body fat mass index. Nutrition 19(7-8):597–604, 12831945.
- 41.↵NorganNG (1994) Population differences in body composition in relation to the body mass index. Eur J Clin Nutr 48 Suppl 3:S10–S25; discussion S26-S27, 7843146.
- 42.↵MeeuwsenSHorganGWEliaM (2010) The relationship between BMI and percent body fat, measured by bioelectrical impedance, in a large adult sample is curvilinear and influenced by age and sex. Clin Nutr 29(5):560–566, 20359792.
- 43.↵JacksonASStanforthPRGagnonJ (2002) The effect of sex, age and race on estimating percentage body fat from body mass index: The Heritage Family Study. Int J Obes Relat Metab Disord 26(6):789–796, 12037649.
- 44.↵GallagherDVisserMSepúlvedaDPiersonRNHarrisTHeymsfieldSB (1996) How useful is body mass index for comparison of body fatness across age, sex, and ethnic groups. Am J Epidemiol 143(3):228–239, 8561156.
- 45.↵MovsesyanLTankóLBLarsenPJChristiansenCSvendsenOL (2003) Variations in percentage of body fat within different BMI groups in young, middle-aged and old women. Clin Physiol Funct Imaging 23(3):130–133, 12752553.
- 46.↵DeurenbergPDeurenberg-YapM (2001) Differences in body-composition assumptions across ethnic groups: practical consequences. Curr Opin Clin Nutr Metab Care 4(5):377–383, 11568498.
- 47.↵GurriciSHartriyantiYHautvastJGDeurenbergP (1998) Relationship between body fat and body mass index: differences between Indonesians and Dutch Caucasians. Eur J Clin Nutr 52(11):779–783, 9846588.
- 48.↵DeurenbergPAndreoliABorgP (2001) The validity of predicted body fat percentage from body mass index and from impedance in samples of five European populations. Eur J Clin Nutr 55(11):973–979, 11641746.
- 49.↵AmaniR (2007) Comparison between bioelectrical impedance analysis and body mass index methods in determination of obesity prevalence in Ahvazi women. Eur J Clin Nutr 61(4):478–482, 17063145.
- 50.↵BozkirliEErtorerMEBakinerOTutuncuNBDemiragNG (2007) The validity of the World Health Organisation's obesity body mass index criteria in a Turkish population: a hospital-based study. Asia Pac J Clin Nutr 16(3):443–447, 17704025.
- 51.↵GaskinPSWalkerSP (2003) Obesity in a cohort of black Jamaican children as estimated by BMI and other indices of adiposity. Eur J Clin Nutr 57(3):420–426, 12627178.
- 52.↵KagawaMKerrDUchidaHBinnsCW (2006) Differences in the relationship between BMI and percentage body fat between Japanese and Australian-Caucasian young men. Br J Nutr 95(5):1002–1027, 16611393.
- 53.↵SampeiMANovoNFJulianoYSigulemDM (2001) Comparison of the body mass index to other methods of body fat evaluation in ethnic Japanese and Caucasian adolescent girls. Int J Obes Relat Metab Disord 25(3):400–408, 11319639.
- 54.↵WHO Expert Consultation (2004) Appropriate body-mass index for Asian populations and its implications for policy and intervention strategies. Lancet 363(9403):157–163, 14726171, Erratum in: Lancet. 2004;363(9412):902.
- 55.↵Deurenberg-YapMChewSKDeurenbergP (2002) Elevated body fat percentage and cardiovascular risks at low body mass index levels among Singaporean Chinese, Malays and Indians. Obes Rev 3(3):209–215, 12164474.
- 56.↵WalkerSPRimmEBAscherioAKawachiIStampferMJWillettWC (1996) Body size and fat distribution as predictors of stroke among US men. Am J Epidemiol 144(12):1143–1150, 8956626.
- 57.↵TsaiCJLeitzmannMFWillettWCGiovannucciEL (2004) Prospective study of abdominal adiposity and gallstone disease in US men. Am J Clin Nutr 80(1):38–44, 15213025.
- 58.↵DurenDLSherwoodRJCzerwinskiSA (2008) Body composition methods: comparisons and interpretation. J Diabetes Sci Technol 2(6):1139–1146, 19885303.
- 59.↵TataranniPARavussinE (1995) Use of dual-energy X-ray absorptiometry in obese individuals. Am J Clin Nutr 62(4):730–734, 7572700.
- 60.↵FieldsDAHunterGR (2004) Monitoring body fat in the elderly: application of air-displacement plethysmography. Curr Opin Clin Nutr Metab Care 7(1):11–14, 15090897.
- 61.↵PiersLSSoaresMJFrandsenSLO'DeaK (2000) Indirect estimates of body composition are useful for groups but unreliable in individuals. Int J Obes Relat Metab Disord 24(9):1145–1152, 11033983.
- 62.↵NuñezCKoveraAJPietrobelliA (1999) Body composition in children and adults by air displacement plethysmography. Eur J Clin Nutr 53(5):382–387, 10369494.
- 63.↵JacksonASPollockMLGravesJEMaharMT (1988) Reliability and validity of bioelectrical impedance in determining body composition. J Appl Physiol 64(2):529–534, 3372410.
- 64.↵MulliePVansantGHulensMClarysPDegraveE (2008) Evaluation of body fat estimated from body mass index and impedance in Belgian male military candidates: comparing two methods for estimating body composition. Mil Med 173(3):266–270, 18419029.
- 65.↵CarrascoFReyesERimlerORiosF (2004) [Predictive accuracy of body mass index in estimating body fatness measured by bioelectrical impedance]. Arch Latinoam Nutr 54(3):280–286, 15807202, Spanish.
- 66.↵ClarosGHullHRFieldsDA (2005) Comparison of air displacement plethysmography to hydrostatic weighing for estimating total body density in children. BMC Pediatr 5:37, 16153297.
- 67.↵DemuraSSatoSKitabayashiT (2006) Estimation of body density based on hydrostatic weighing without head submersion in young Japanese adults. J Sports Sci 24(6):589–596, 16608771.
- 68.↵DeurenbergPYapMvan StaverenWA (1998) Body mass index and percent body fat: a meta analysis among different ethnic groups. Int J Obes Relat Metab Disord 22(12):1164–1171, 9877251.
- 69.↵JacksonASPollockMLWardA (1980) Generalized equations for predicting body density of women. Med Sci Sports Exerc 12(3):175–181, 7402053.
- 70.↵DeurenbergPvan der KooyKLeenenRWeststrateJASeidellJC (1991) Sex and age specific prediction formulas for estimating body composition from bioelectrical impedance: a cross-validation study. Int J Obes 15(1):17–25, 2010255.
